The proposed project is designed to enrich our understanding of the nature and development of speed of information processing in infancy and to explore its relation to working memory. This work will provide new understanding of the early development of speed of information processing, one of the core constructs of infant cognition and possibly one of the foundations of continuity between cognition in infancy and later childhood. The principal aims are: (1) to determine the extent to which attention, affect, and inhibitory ability are related to speed of visual information processing; (2) to examine the relation of speed of visual information processing to the encoding of configural and featural information; (3) to characterize the developmental trajectory of speed of visual information processing over the course of the first year of life; (4) to similarly characterize the developmental trajectory of working memory over the same period; and (5) to examine the relation of speed of visual information processing to working memory, and the extent to which age-related changes in speed account for age-related changes in memory. These issues will be examined in the first year of life for two distinct and conceptually important domains of visual processing: faces and objects. Aims 1 and 2 will involve parallel studies for faces and objects at each of three ages (5, 7, & 9 months) in which speed of processing will be assessed with a new task, continuous familiarization. This task involves repeated presentations of a familiar stimulus, each time paired with a novel one; speed is indexed by the time taken to show a consistent preference for the novel. Attention and affect will be assessed concurrently, as will the infant's propensity to inhibit looking and shift gaze. Aims 3 - 5 will be investigated both cross-sectionally and longitudinally. Infants aged 3-13 months will be tested with the continuous familiarization task and a span task measuring working memory. Developmental curves will be compared for cross-sectional and longitudinal samples and across domains (faces and objects).